World Equity Markets and COVID-19: Immediate Response and Recovery Prospects - Munich Personal ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Munich Personal RePEc Archive World Equity Markets and COVID-19: Immediate Response and Recovery Prospects Seven, Unal and Yilmaz, Fatih Central Bank of the Republic of Turkey, King Abdullah Petroleum Studies and Research Center 8 June 2020 Online at https://mpra.ub.uni-muenchen.de/100987/ MPRA Paper No. 100987, posted 15 Jun 2020 04:24 UTC
World Equity Markets and COVID-19: Immediate
Response and Recovery Prospects∗†
Ünal Seven Fatih Yılmaz
Central Bank of King Abdullah Petroleum Studies
the Republic of Turkey and Research Center
June 8, 2020
Abstract
The actual macroeconomic impacts of the COVID-19 pandemic will be realized over time;
however, its impact on financial markets was much faster and dramatic. Following the spread
of the pandemic, most global equity markets experienced significant falls and started to re-
bound with the announcement of economic rescue packages. However, the equity markets’
responses to the packages have varied across countries. In this paper, we first look at what
may explain the differences in the equity market falls across-countries. Secondly, we study the
systematic relation between the size and the type of rescue packages, the severity of the out-
break and the recovery performance of equity markets. Using cross-country OLS regressions,
we find that investors’ immediate reaction to equity markets in countries with higher pan-
demic related deaths is more negative. Moreover, our results show that not all types of rescue
packages are effective in restoring investors’ valuation of equity markets. Among different
types, fiscal stimulus support seems to be a stronger predictor of equity market recovery per-
formance. We also find that the severity of the outbreak in each country affected the equity
markets’ reactions.
Keywords: COVID-19, Stock Markets, Stimulus Packages, Recovery
JEL Classifications: G15, G18
∗ The views in this paper are solely the responsibilities of the authors and should not be interpreted as reflecting
the view of their affiliated institutions.
† E-mail addresses: unal.seven@tcmb.gov.tr, fatih.yilmaz@kapsarc.org.World Equity Markets and COVID-19
1 Introduction
The COVID-19 pandemic, initially broke out in China in early January, has since posed a severe
threat to our lives and economies. Millions of infections and hundreds of thousands of deaths
have been recorded globally, with more casualties expected in the coming days and months. In
response, many countries have had to halt daily life and suspend all but essential international
travel. The measures1 taken to limit the spread of COVID-19 have hit all economies. While
the actual macroeconomic impacts of the pandemic will be realized over time, its impact on
financial markets was much faster and dramatic.2 Most major global equity markets experienced
significant falls up to 50%. Recognizing the short and long-run economic impacts of COVID-19,
many governments announced multi-trillion US Dollars economic rescue packages since mid-
March. Following this positive signal, global equity markets started to rebound with gradual
recovery and declining volatility. With this background, using benchmark equity market indices,
we raise two important questions in this paper. Firstly, we ask what may explain the differences
in equity market falls across-countries? And secondly, is there a systematic relation between the
size and type of rescue packages, the pandemic related death rate and the recovery performance
of equity markets? The first question lays down the main factors behind the investors’ immediate
reaction to the COVID-19 pandemic, while the second one will inform us more about the future
prospects.
In our analysis, we use equity market benchmark indices from 78 countries and google
search data. We first identify the key dates for the impact of the pandemic on world equity
markets by using the announcement of economic rescue packages and google search data on the
coronavirus. Most world equity markets seem to have reached to their peaks around February
19 that corresponds to the early days of the pandemic spreading to Italy. Similarly, most equity
market indices have reached to their dips around March 23, when many developed economies
including US and European countries started to declare rescue packages. Using these dates,
we compute loss rate, the percentage difference between the peak and dip performances of the
benchmark indices, and the recovery rate, percentage of loss that has been recovered until the
end of April. We then conduct a cross-country empirical analysis on explaining the differences
in the loss and recovery rates based on country level macroeconomic variables, including recent
COVID-19 bail outs.
With regard the first question, we find that investors’ immediate reaction to equity markets
in countries with higher pandemic related deaths appears to be more negative. This is actually
in line with our prior expectations. In the early days of the pandemic, uncertainty was unprece-
dented, while the pandemic related cases and deaths seemed to be the only and timely available
information. Our results also show close and statistically significant correlation between the
market loss rate and median age of the domestic population, indicating that countries with older
1 Hale et al. (2020) introduced a stringency index to combine measures and describe variation in government
responses in various areas.
2 Zhang et al. (2020) showed that risks associated to the pandemic caused to higher volatility in global financial
markets.
1World Equity Markets and COVID-19
populations experienced higher losses in the equity markets. With regard to the second question,
we find that not all type of rescue packages are effective in restoring investors valuation of equity
markets. In particular, fiscal stimulus policies among others (e.g., easing financial market regula-
tions, rate cuts, etc.,) seem to be strongly correlated with higher recovery in the equity markets.
Since the pandemic related risks increased the volatility in global financial markets (Zhang et
al., 2020) one would expect to see a negative impact of the pandemic related death rate on the
equity market recovery. Therefore, we also control for the COVID-19 related death rate in the
regressions. As expected, we observe a negative significant relationship between the death rate
and recovery rate. Due to general expectations of slow recovery in tourism sector (i.e., due to
global lockdown) and commodity markets (i.e., especially oil and gas markets due to drop in
production and consumption), and also, disruption on global value chains, we include tourism
revenues, natural resources richness and export revenues of countries to the main model.3 While
the initial findings preserved, we also show that higher tourism revenues or resource richness is
associated with slower recovery in global equity markets. However, we do not find any statis-
tically significant relation between export revenues of a country and itsequityk market recovery
performance in the short time period analyzed in our study.
Our results are consistent with the newly emerging literature. For example, Zhang et al.
(2020) provided a statistical analysis of the impact of COVID-19 pandemic on equity market
risks. They found that financial market risks have increased in response to the pandemic, and
the severity of the outbreak in each country affected the equity markets’ reactions. Similarly,
Zaremba et al. (2020) studied whether government interventions against the spread of pandemic
affect equity market volatility by using data of 67 countries. Using daily equity market data for
the period from 1 January 2020 to 3 April 2020 and simple panel data regressions, they showed
that government responses increase the volatility in international equity markets. They, how-
ever, did not look at the relationship between government responses and equity market returns.
Moreover, using data on the stocks of Hang Seng Index and Shanghai Stock Exchange Compos-
ite Index (SSE), Al-Awadhi et al. (2020) found a negative significant correlation between equity
returns, the daily growth in total confirmed cases and the daily growth in total pandemic related
deaths. Although their work focused on the equity market return-pandemic case relationship,
their results come from a single country sample, unlike ours. Therefore, our study can fill these
gaps by providing a cross-country evidence on both the loss rate and the recovery rate. However,
we should note that our results do not suggest any causality. We are very well aware of the
fact that our regression results may suffer from endogeneity and omitted variable bias problems,
for instance, the type of rescue plan and the equity market response are likely co-determined
by an omitted variable. Despite these limitations, our study provides suggestive evidence from
3 Tourism is perhaps the most affected sector due to lockdowns and halting of international travel. According
to the OECD (2020), the pandemic related shocks could decrease the international tourism economy by 45-70% in
2020. Similarly, commodity markets lost significant value with very volatile prices which creates concerns on the
future of resource rich economies, and commodity-dependent emerging and developing economies are among the
most vulnerable economies to COVID-19 since their exports and fiscal revenues are substantially declining (UN, 2020;
World Bank, 2020). Lastly, disruption on domestic and global trade network is another widely discussed impact of
the pandemic (see WTO, 2020, for the export prohibitions and restrictions).
2World Equity Markets and COVID-19
strong correlations without claiming any causal inference and contributes to the discussions on
understanding the immediate impact of the pandemic on financial markets and how the policy
responses may restore investors’ confidence on markets.
The next section presents the details of our data and also, discusses the economic timeline
of the pandemic. Section 3 presents our econometric results. Finally, Secion 4 concludes.
2 Data
Our main data, benchmark equity market indices of 78 countries, is collected from Bloomberg.4
The data is daily and covers the period of January 1 - April 25, 2020. It contains all the major
economies from all the regions of the world. Using this data and Google search data, we identify
the reference dates for when most of the markets have peaked and then, recorded dips. We use
the peaks and dips to measure the impact of the COVID-19 on equity markets (i.e., loss rate) and
how much of this loss recovered (i.e., recovery rate). We explain each concept step-by-step below.
How did COVID-19 Spread Over the World Stock Markets?
Figure 1 presents the market indices, which we normalized to be between zero and one.5 In
Figure 1a, we present a selection of countries from different continents to show the spread of
COVID-19 globally. Three key dates – January 13, February 19 and March 23 – flash out from the
figure. The first date shows the date that China’s SSE peaked. The second date shows the peak
for most of the countries in the rest of the world, and finally, the third date is the beginning of
recovery process. We show the performance of all the equity market benchmark indices in our
full sample during these dates in Figure 1b. The general tendency appears to be preserved in the
full sample.
In the early days of the COVID-19 pandemic, the focus was primarily on China and the
expectations of potentially mild effects of the virus on its economy. However, with the the spread
of the virus to Europe in mid-February, it started to become a global phenomenon, as the focus
turned to the potential for it to precipitate a significant global economic slump. In Figures 2a
and 2b, we show the number of countries peaked in January and February, and recorded dips
in March. According to the first figure, some moderate number of countries, including Asian
countries surrounding China and major oil exporters, reached to their peaks around the same
time with China, while many more countries in the rest of the world peaked around February
19. Finally, most countries (63 % of our sample) reached to their dips between March 18 and 23.
The timeline implied by the performance of world equity markets seem to be closely inline
with Google search statistics for the keyword “coronavirus” in the world and in major countries.
4 Very small islands, countries without a equity market or a benchmark index, equity markets that were closed
during the pandemic period or related data not provided by Bloomberg are excluded from the study. We also could
not include some countries due to availability on other variables used in the analysis. Due to these data limitations,
we are left with 78 countries. A full list of countries, covered in our study, can be found in Table A.1 in the Appendix.
5 The normalization is done by [x-min(x)]/[min(x)-max(x)], where x is index value.
3World Equity Markets and COVID-19
Figure 1: Stock Exchange Performances (Normalized)
a) Major Stock Exchange Indices b) All Stock Exchange Indices
1
1
March 23
.8
March 23
.8
.6
.6
January 13
.4
February 19
.4
.2
January 13 February 19
0
.2
01jan2020 15jan2020 29jan2020 12feb2020 26feb2020 11mar2020 25mar2020 08apr2020 22apr2020
Mean (Sample) China United States United Kingdom
Germany Italy Japan Australia
0
Brazil India
01jan2020 15jan2020 29jan2020 12feb2020 26feb2020 11mar2020 25mar2020 08apr2020 22apr2020
Source: Bloomberg.
Note: All indexes are normalized to [0,1] interval.
Figure 2: Identifying Stock Markets’ Peaks and Dips
a) Number of Country Peaks Over Time b) Number of Country Dips Over Time
14
25
12
20
10
# of Dipping Country
# of Peaking Country
15
6 8
10
4
5
2
0
0
15jan2020 22jan2020 29jan2020 05feb2020 12feb2020 19feb2020 26feb2020 04mar2020 11mar2020 18mar2020 25mar2020 01apr2020
Source: Authors’ calculations from the Bloomberg data.
In Figure 3, we show Google search statistics for the World and China (i.e., the origin of the
pandemic), Italy (i.e., the first infected in Europe), and finally, the United States (i.e., a major
economy). It is interesting to see that in the early day of the pandemic, there was little interest
on the issue in the world, perhaps due to the assumption that the pandemic would be preserved
locally. However, the interest on the pandemic quickly picked up around the same time that
4World Equity Markets and COVID-19
the virus spread to Italy and then, the world equity markets immediately responded. As we
reached to mid-March, the interest started to declining. To fight the negative economic impact
of COVID-19, many governments started to announce unprecedented economic rescue packages
since mid-March.6 Some of the most significant ones include a 2 trillion stimulus package in the
United States (U.S.) (10% of its GDP) and a 0.8 trillion package in Germany (20.5 % of its GDP).
The giant rescue packages appear to restore investor confidence, given that the markets started
pick up since March 23. With this background, we treat February 17–21, as the global peak, and
March 18-23, as the global dip, and use these dates to compute equity market loss and recovery
rates for all the countries in our sample.
Figure 3: Google Search on Coronavirus
100
March 18-23
February 17-21
80
January 13
60
40
20
0
01jan2020 22jan2020 12feb2020 04mar2020 25mar2020 15apr2020
World China
Italy United States
Source: Google search statistics.
Measuring the Stock Market Loss due to COVID-19
In computing the loss rate, we take the maximum index value in the global peak days to deter-
mine each market’s peak and similarly, take the minimum index value in the global dip days to
determine each market’s dip. Using the percentage difference between dip and peak values, we
measure the loss rate for each country as the immediate impact of the the COVID-19 on their
equity markets. Figure 4 displays the computed loss rate for each country in our sample. The
figure shows that the pandemic crashed world equity markets up to 50% in some countries. Aus-
trian ATX Prime index dropped for 48%, Italian MIB index declined for 41%, German DAX index
fell for 39%, US Dow Jones dropped for 37%, and the UK FTSE100 index was down for 33%.
6 Most of these packages were first communicated with the public usually a few days before the actual legal process
in respective parliaments and congresses. For instance, the initial Coronavirus Aid, Relief, and Economic Security Act
in the United States was designed to be 1.4 trillion USD, which did not pass the Senate in March 23. The revised bill
with 2 trillion USD passed in March 25. However, global markets were already informed about the bailout on March
24. For the full story, see the Washington Post article.
5World Equity Markets and COVID-19
Figure 4: Loss Rate Between the Peak and Dip
0
-10
Loss Rate (%)
-30 -40
-50 -20
Ghana
Mongolia
Slovak Republic
Kazakhstan
China
Laos
Qatar
Tunisia
Montenegro
Oman
Cambodia
Latvia
Bahrain
Malaysia
Malta
Nigeria
Lithuania
Hong Kong
Kenya
Iceland
Jamaica
Saudi Arabia
Serbia
Denmark
Mexico
Switzerland
Bulgaria
Estonia
Morocco
Kuwait
Peru
Pakistan
New Zealand
Japan
Vietnam
North Macedonia
Turkey
Slovenia
Singapore
Cyprus
Romania
Sweden
Israel
Poland
Croatia
United Kingdom
United Arab Emirates
Norway
Indonesia
Portugal
Thailand
South Korea
Netherlands
Finland
Hungary
Australia
Egypt
United States
India
Canada
Czech Republic
Philippines
Chile
Spain
France
Belgium
Germany
Ireland
Italy
Namibia
Luxembourg
Argentina
South Africa
Brazil
Russia
Colombia
Greece
Austria
Source: Authors’ calculations from the Bloomberg data.
Note: Loss rate is computed as the percentage change between dip and peak values of each index.
Measuring the Stock Market Recovery Rate
We measure the recovery rate for a equity market benchmark index (index) as the ratio of the
recovered amount (from its dip to April 24, the last data in our sample) to the loss amount,
between the peak and dip dates.
index(last) − index(dip)
recovery rate = (1)
index(dip) − index( peak)
For convenience, we multiple the denominator with negative one to make the loss rate positive
for every index.7 Hence, the recovery becomes positive, when there is actually a positive recov-
ery; zero, if no recovery observed, and negative, if further losses accrued. Despite the growing
concerns of a potential second wave of the pandemic, following the stimulus policies, equity
markets have made significant recoveries since their dips. As of the last day in our sample, the
US Dow Jones had recovered 48% of the losses it sustained between February 19 and March 23.
7 All the equity market indices in our sample recorded losses during this time and thus, shows with positive
numbers.
6World Equity Markets and COVID-19
During the same period, the German DAX index recovered 35%, and the UK FTSE 100 regained
31% of its losses.
Figure 5: Recovery Rate from the Dip
75
50 25
Recovery Rate (%)
-50 -25 -75
-1000
Mongolia
Ghana
Bahrain
Oman
Morocco
Peru
Montenegro
Cyprus
Spain
Jamaica
Mexico
Nigeria
Italy
Qatar
Laos
Luxembourg
Slovak Republic
Hungary
Kuwait
Tunisia
Brazil
Belgium
Australia
Egypt
France
Greece
Bulgaria
Romania
Indonesia
Portugal
Singapore
Colombia
Estonia
South Africa
Slovenia
Austria
Kenya
Croatia
Poland
Philippines
Saudi Arabia
United Kingdom
Malta
Ireland
Serbia
Hong Kong
Finland
India
Israel
Russia
Norway
Pakistan
Germany
Sweden
Japan
Czech Republic
China
North Macedonia
United Arab Emirates
Turkey
Namibia
Cambodia
Argentina
Vietnam
Iceland
Netherlands
Chile
Thailand
Switzerland
Malaysia
Canada
United States
New Zealand
South Korea
Kazakhstan
Lithuania
Denmark
Latvia
Source: Authors’ calculations from the Bloomberg data.
Note: Recovery rate is computed as the ratio of the recovered amount from the dip of each index to its value on April 24 to the loss amount
between the peak and dip dates.
COVID-19 Rescue Measures and Other Variables
Using the IMF COVID-19 Policy Tracker data8 , Elgin et al. (2020) created a COVID-19 Economic
Stimulus Index (CESI) that we use to measure the size of rescue packages in the analysis. The res-
cue packages contain various types of supports for households, firms, and the financial industry.
Fiscal supports usually occupy the biggest portion in packages, which include direct cash trans-
fers (e.g., enhanced employment benefits, improved food safety, etc.), tax rebates (to individuals
and firms), credit guarantees (i.e., preventing firm bankruptcies), forgiving loans (i.e., usually
for small and medium size enterprises). Monetary and macro-financial supports usually include
easing market liquidity with policy rate cuts and relaxing regulations and supervision controls,
as well as asset purchase programs by the central banks. We use both the composite index and
8 The IMF produces an up-to-date policy tracker database that summarizes government responses against the
human and economic impact of the pandemic for 193 economies, see the IMF Policy Tracker Database.
7World Equity Markets and COVID-19
its components in our analysis. The summary statistics for all the variables used in our analysis
are presented in Table 1 and their cross-correlations are shown in the Appendix, Table A.2.
Table 1: Summary Statistics
The table presents the summary statistics for all the variables used in the analysis. Loss rate is computed as the percentage change between
dip and peak values of each index while recovery rate is computed as the ratio of the recovered amount from the dip of each index to
its value on April 24 to the loss amount between the peak and dip dates. Median age is the median age of the population for the year
2019. Hospital beds is the number of hospital beds per 1000 people. The CESI index is a composite index for measuring the combined
impact of all adopted policies against the economic impact of the COVID-19 pandemic measures as defined by Elgin et al. (2020), and
it is normalized to [0, 1] interval. Fiscal stands for the fiscal policy package as a percent of GDP, Rate Cut is the interest rate cut as a
percent of the pre-crisis level, Macro-Financial is the monetary stimulus package as a percent of GDP, BoP is the monetary intervention to
control the balance of payments and the exchange rate as a percent of GDP and finally, Other BoP is a dummy variable taking the value
of 1 if there are other accompanying measures towards stabilizing BoP and exchange rate a lâ Elgin et al. (2020). Resource rich takes
value of 1 for resource rich countries, zero otherwise. Tourism revenue represents the international tourism receipts as a percentage of GDP
while total exports variable is defined as the ratio of total exports of goods and services to GDP. Finally, population is the total number of people.
Variable Obs Mean Std. Dev. Min Max
Loss Rate (%) 78 29.95 10.02 2.23 48.05
Recovery Rate (%) 78 28.46 23.35 -99.07 72.18
Median Age 78 36.54 7.40 18 48
Death Rate (per 1,000,000 people) 78 3.16 12.61 0 100.56
GDP per capita (000 USD) 78 26.16 23.89 1.20 110.74
Hospital Beds (per 1,000 people) 78 3.81 2.49 0.50 13.40
CESI Index 78 0.37 0.27 0 1
Fiscal (% of GDP) 78 6.01 5.34 -5 22.14
Rate Cut (%) 78 22.42 32.16 -105.56 100
Reserve Requirement and Buffer (%) 78 25.05 34.76 -8.33 100
Macro-Financial (% of GDP) 78 5.07 6.38 0 28
BoP (% of GDP) 78 0.26 0.98 0 6.50
Other BoP 78 0.26 0.44 0 1
Resource Rich Dummy 78 0.22 0.42 0 1
Tourism Revenue (% of GDP) 78 4.93 4.89 0.27 22.89
Total Exports (% of GDP) 78 52.57 39.02 8.24 223.08
Population 78 76300000 221000000 353574 1390000000
3 Empirical Analysis
In this section, we present two sets of results. The first set shows our findings on what may
explain the cross-country differences in their loss rates and the second set displays our results on
explaining the cross-country differences in their recovery rates.
3.1 What Explains the Cross Country Differences in Loss Rates?
We first provide scatter plots for certain country characteristics and the loss rate. Panel (a) of
the Figure 6 shows the cross country relationship between the total number of COVID-19 related
deaths (up to March 24, 2020)9 and the loss rate. Strong positive relationship between the two
9 We
take the total number of pandemic related deaths up to March 24, 2020 since we consider the period of March
18-23 as the global dip for equity markets.
8World Equity Markets and COVID-19
variables is evident. We observe a similar relationship between GDP per capita and the loss rate,
which could be somehow surprising. However, we should note that most advanced economies
implemented more strict policies (e.g., social distancing, shutdowns, etc.) to prevent the spread
of the pandemic. This could explain the strong positive correlation between the two variables
such that tighter measures may imply weakening of firms’ future profits from the perspective of
investors.
Figure 6: Loss Rate, Number of Deaths and GDP per capita
a) Loss Rate and Number of Deaths b) Loss Rate and GDP per capita
50
50
40
40
Loss Rate (%)
Loss Rate (%)
30
30
20
20
10
10
0
0
0 2 4 6 8 7 8 9 10 11 12
Log (Number of Deaths) Log (GDP per capita)
Loss Rate Fitted values Loss Rate Fitted values
Note: Number of deaths represents the total number of COVID-19 related deaths till March 24, 2020. Loss rate is computed as the percentage
change between dip and peak values of each index.
We then conduct simple cross country OLS regressions with the loss rate as our dependent
variable. We control for the main factors that could explain the cross country differences in
loss rates, including, median age of the population, death rate (per a million population), GDP
per capita and number of hospital beds.10 In doing so, we aim to control for the pandemic,
demographic, income and health system related differences across countries.
Table 2 presents the regression results by adding one variable to the baseline model in each
column. The results show statistically significant and positive correlation between the loss rate
and median age of population, pandemic related death rate, and the GDP per capita. Median
age only becomes statistically insignificant in the last column with almost no change on sign
and magnitude. Pandemic related death rate also remains to be statistically significant across
the table. Hospital beds (per 1,000 people), as a proxy for the health system capacity to fight
the pandemic, is not statistically significant. However, its sign is expectedly negative, implying
a negative relation with the loss rate. These results imply that countries with relatively older
populations and higher pandemic related death rates experienced a larger immediate drop in
the equity markets following the spread of the virus to Europe. The results should not come at a
10 Given the degrees of freedom concerns with such a small sample, we only included the main variables.
9World Equity Markets and COVID-19
big surprise, as there was huge uncertainty in the early days of the pandemic and investors did
not have much other information to evaluate the impact of the pandemic on markets. As pointed
out earlier, observing a higher loss in richer countries may be due to the fact that richer countries
took faster action in halting the daily life.
Table 2: Cross-Country Regression on the Loss Rate
The table presents the cross-country OLS results for the regression between the loss rate and certain country
characteristics. The dependent variable, loss rate, is computed as the percentage change between dip and peak values
of each equity market index. Median age represents the median age of the population in 2019. Death rate is the ratio
(per 1 million people) of total number of COVID-19 related deaths till March 24 to population. Hospital beds is the
total number of hospital beds per 1,000 people. Robust (clustered) standard errors in parentheses. The clustering is
done at the World Bank eight region level. *** pWorld Equity Markets and COVID-19
Figure 7: Recovery Rate and Stimulus Packages
a) Recovery Rate and CESI Index b) Recovery Rate and Fiscal Package
100
100
50
50
Recovery Rate (%)
Recovery Rate (%)
0
0
-50
-50
-100
-100
0 .2 .4 .6 .8 1 -5 0 5 10 15 20
CESI Index Fiscal Package (% of GDP)
Recovery Rate Fitted values Recovery Rate Fitted values
Notes: The CESI index is normalized to [0,1] interval. Recovery rate is computed as the ratio of the recovered amount from the dip of each index
to its value on April 24 to the loss amount between the peak and dip dates.
include controls like resource richness dummy, tourism and total export revenues as a percentage
of GDP. The results show that there is a positive but statistically insignificant relationship between
recovery rate and the CESI index (column 1 of Table 3), and this result does not change when
we control for country specific variables (columns 2 and 3 of Table 3). In columns 4, 5 and 6, we
replace the main index with its sub-components. Among the six sub-components of CESI index,
the fiscal, reserve requirement and buffer packages are positively and statistically significantly
correlated with recovery rate. The fiscal stimulus packages seem to be more strongly correlated
with the recovery rate than other sub-components. However, we do not observe any significant
relationship between rate cut and recovery rate while also macro-financial policy package and
other BoP variable have insignificant relationships with recovery rate.
Besides the main results, we also find a negative significant relationship between the COVID-
19 related death rate and recovery rate, suggesting that countries with high pandemic related
death rate tend to experience lower recovery rate in the equity market. We also observe statisti-
cally significant relation between the recovery rates and countries resource richness, and tourism
revenues. This is to say concerns on energy demand and swinging oil and gas prices had a severe
impact on the equity market performances of resource rich economies.11 Similarly, halting daily
life and imposing restrictions on international activity also negatively affected the economies that
rely more on tourism revenues. These general concerns seem to be reflected on the equity market
performances according to our results. Amid the discussion on a potential shrink of world trade,
export variable does not seem to be statistically significantly related to recovery rate.
11 For
a more detailed discussion on COVID-19 impact on oil and gas exporting economies and different type of
green versus traditional assets, see Yilmaz (2020).
11World Equity Markets and COVID-19
Table 3: Cross-Country Regressions on the Recovery Rate
The table presents the cross-country OLS results for the regression between the recovery rate, the pandemic related death rate, economic stimulus
packages and certain country characteristics. The dependent variable, recovery rate, is computed as the ratio of the recovered amount from the dip
of each equity market index to its value on April 24 to the loss amount between the peak and dip dates. The CESI index is a composite index for
measuring the combined impact of all adopted policies against the economic impact of the COVID-19 pandemic measures as defined by Elgin et
al. (2020), and it is normalized to [0, 1] interval. Fiscal stands for the fiscal policy package as a percent of GDP, Rate Cut is the interest rate cut
as a percent of the pre-crisis level, Macro-Financial is the monetary stimulus package as a percent of GDP, BoP is the monetary intervention to
control the balance of payments and the exchange rate as a percent of GDP and finally, Other BoP is a dummy variable taking the value of 1 if
there are other accompanying measures towards stabilizing BoP and exchange rate a lâ Elgin et al. (2020). Death rate is the ratio (per 1 million
people) of total number of COVID-19 related deaths till March 24 to population. Resource Rich takes value of 1 for resource rich countries, zero
otherwise. Tourism revenue represents the international tourism receipts as a percentage of GDP while total exports variable is defined as the
ratio of total exports of goods and services to GDP. Robust (clustered) standard errors in parentheses. The clustering is done at the World Bank
eight region level. *** pWorld Equity Markets and COVID-19
to restore expectations. In particular, countries with larger fiscal rescue packages seem to have
experienced a higher recovery. Our findings show that there is a significant negative relation-
ship between the COVID-19 related death rate and recovery rate, suggesting that the severity of
the pandemic in each economy affects the equity market’s reaction. We also show that countries
with high dependence on natural resources and tourism revenues appear to experience a slowing
down in their recovery rates.
We should note that our study provides suggestive evidence from strong correlations with-
out claiming any causal inference, as it may have limitations, such as omitted variable bias. How-
ever, the presented results contribute to the discussions on understanding the immediate impact
of the pandemic on financial markets and how the policy responses may restore investors’ confi-
dence on markets. This may also provide some inference for the prospective recovery of financial
markets from the COVID-19 pandemic. Another limitation of our study can be the narrow re-
search sample, which is mostly due to the data availability. We will be able to update and verify
our findings as richer data sets become available.
References
[1] Adam, Zaremba , Renatas, Kizys , David Y., Aharon , Ender, Demir. 2020. “Infected Mar-
kets: Novel Coronavirus, Government Interventions, and Stock Return Volatility around the
Globe’.” Finance Research Letters. (doi: https://doi.org/10.1016/j.frl.2020.101597)
[2] Al-Awadhi, Abdullah M., Alsaifi, Khaled, Al-Awadhi, Ahmad, Alhammadi, Salah . 2020.
“Death and Contagious Infectious Diseases: Impact of the COVID-19 Virus on Stock Market
Returns.” Journal of Behavioral and Experimental Finance 27. 100326.
[3] Baldwin, Richard, Beatrice Weder di Mauro. 2020. “Mitigating the COVID Economic Crisis:
Act Fast and Do Whatever It takes.” VoxEU.org eBook, CEPR Press.
[4] Elgin, Ceyhun, Basbug, Gokce, Yalaman, Abdullah. 2020. "Economic Policy Responses to a
Pandemic: Developing the COVID-19 Economic Stimulus Index" COVID Economics: Vetted
and Real Time Papers 3: 40-54.
[5] Gourinchas, Pierre-Olivier. 2020. "Flattening the Pandemic and Recession Curves." in Miti-
gating the COVID Economic Crisis: Act Fast and Do Whatever. Edited By Richard Baldwin
and Beatrice Weder di Mauro. London: CEPR Press.
[6] Hale, Thomas, Noam Angrist, Beatriz Kira, Anna Petherick, Toby Phillips, Samuel Webster.
2020. “Variation in Government Responses to COVID-19.” Version 5.0. Blavatnik School of
Government Working Paper.
[7] Kinda, T., Mlachila, M., and Ouedraogo, R. 2018. “Do Commodity Price Shocks Weaken the
Financial Sector?” The World Economy, 41(11), 3001-3044.
13World Equity Markets and COVID-19
[8] OECD. 2020. “Tackling Coronavirus (COVID-19): Tourism Policy Responses.” Technical Re-
port.
[9] Tokhir N Mirzoev, Ling Zhu, Yang Yang, Tian Zhang, Erik Roos, Andrea Pescatori and
Akito Matsumoto. 2020. "The Future of Oil and Fiscal Sustainability in the GCC Region."
IMF Departmental Papers/Policy Papers 20/01.
[10] United Nations. 2020. “Commodity Exporters Face Mounting Economic Challenges as Pan-
demic Spreads”. Policy Brief No:60.
[11] World Bank. 2020. "Commodity Market Outlook: Implication of COVID-19 for Commodi-
ties". Technical Report.
[12] WTO. 2020. “Export Prohibitions and Restrictions.” Information Note.
[13] Yilmaz, F. 2020 “Long and Short-Term Economic Impact of COVID-19: Does the Future
Look Greener?” KAPSARC Instant Insight, forthcoming .
[14] Zhang, Dayong, Min Hu, and Qiang Ji. 2020. "Financial Markets Under the Global Pandemic
of COVID-19." Finance Research Letters, 101528
14World Equity Markets and COVID-19
Appendix
Table A.1: List of Countries Included in the Study
Country Region Country Region
Argentina Latin America and Caribbean Luxembourg Europe and Central Asia
Australia East Asia and Pacific Malaysia East Asia and Pacific
Austria Europe and Central Asia Malta Middle East and North Africa
Bahrain Middle East and North Africa Mexico Latin America and Caribbean
Belgium Europe and Central Asia Mongolia East Asia and Pacific
Brazil Latin America and Caribbean Montenegro Europe and Central Asia
Bulgaria Europe and Central Asia Morocco Middle East and North Africa
Cambodia East Asia and Pacific Namibia Sub-Saharan Africa
Canada North America Netherlands Europe and Central Asia
Chile Latin America and Caribbean New Zealand East Asia and Pacific
China East Asia and Pacific Nigeria Sub-Saharan Africa
Colombia Latin America and Caribbean North Macedonia Europe and Central Asia
Croatia Europe and Central Asia Norway Europe and Central Asia
Cyprus Europe and Central Asia Oman Middle East and North Africa
Czech Republic Europe and Central Asia Pakistan South Asia
Denmark Europe and Central Asia Peru Latin America and Caribbean
Egypt Middle East and North Africa Philippines East Asia and Pacific
Estonia Europe and Central Asia Poland Europe and Central Asia
Finland Europe and Central Asia Portugal Europe and Central Asia
France Europe and Central Asia Qatar Middle East and North Africa
Germany Europe and Central Asia Romania Europe and Central Asia
Ghana Sub-Saharan Africa Russia Europe and Central Asia
Greece Europe and Central Asia Saudi Arabia Middle East and North Africa
Hong Kong East Asia and Pacific Serbia Europe and Central Asia
Hungary Europe and Central Asia Singapore East Asia and Pacific
Iceland Europe and Central Asia Slovak Republic Europe and Central Asia
India South Asia Slovenia Europe and Central Asia
Indonesia East Asia and Pacific South Africa Sub-Saharan Africa
Ireland Europe and Central Asia South Korea East Asia and Pacific
Israel Middle East and North Africa Spain Europe and Central Asia
Italy Europe and Central Asia Sweden Europe and Central Asia
Jamaica Latin America and Caribbean Switzerland Europe and Central Asia
Japan East Asia and Pacific Thailand East Asia and Pacific
Kazakhstan Europe and Central Asia Tunisia Middle East and North Africa
Kenya Sub-Saharan Africa Turkey Europe and Central Asia
Kuwait Middle East and North Africa United Arab Emirates Middle East and North Africa
Laos South East Asia United Kingdom Europe and Central Asia
Latvia Europe and Central Asia United States North America
Lithuania Europe and Central Asia Vietnam East Asia and Pacific
15World Equity Markets and COVID-19
Table A.2: Correlation Coefficients
The table reports the correlation coefficients for the main variables. Loss rate is computed as the percentage change between dip and peak values of
each equity market index. Recovery rate is computed as the ratio of the recovered amount from the dip of each equity market index to its value
on April 24 to the loss amount between the peak and dip dates. The CESI index is a composite index for measuring the combined impact of all
adopted policies against the economic impact of the COVID-19 pandemic measures as defined by Elgin et al. (2020), and it is normalized to [0, 1]
interval. Death rate is the ratio (per 1 million people) of total number of COVID-19 related deaths till March 24 to population. GDP per capita is
the real GDP per 1,000 people. Hospital beds is the total number of hospital beds per 1,000 people. Tourism revenue represents the international
tourism receipts as a percentage of GDP while total exports variable is defined as the ratio of total exports of goods and services to GDP. * pYou can also read
